Non-Relational Databases in Big Data

Abstract

These days' Big data is becoming a very essential component for the industries where large volume of data at very high speed is used to solve particular data problems. Generally, big data is first analyzed and then used with other available data in the company to make it more effective. Therefore, big data is never operated in isolation. There are a variety of non-relational data stores (databases) available. These data stores and big data can be used in combination to work with. Attributes of these databases are available for companies where big data is used. In last few years it is the requirement of companies that these databases should operate very fast, it should be extended/contracted whenever required and should generate reports quickly. It also requires that the different means should be available to manage and organize these massive databases. This paper mainly focuses on some methods for data management like key-value databases, document databases, tabular databases, object data bases and graph databases. Use of RDBMS for big data implementation is not practical because of its performance, scale or even cost. Now a day's companies have adopted non-relational databases, known as NoSQL databases. Programmers and analysts may take benefit of non-relational databases as it has simple modeling constraints than the relational databases. Analysts can do various types of analysis by taking different types of non-relational databases every time. For example, key value databases, graph databases. The non-relational databases do not depend on the common traditional relational database management systems.